Adaptive Learning for Invoice Lines enhances Intelligent Document Recognition (IDR) by learning from user corrections made in the interactive invoice viewer. This feature specifically applies to certain line-level attributes. When a user corrects the Amount, Quantity, or Description on an invoice line, the system's machine learning algorithms are trained. This improves the accuracy of data extraction for these fields on future invoices from the same supplier, leading to increased touchless processing rates. Other correctable line-level attributes not listed as options include Unit Price and UOM (Unit of Measure).

C. Supplier is a header-level attribute, not a line-level attribute. Adaptive learning for suppliers is a separate capability.

E. Invoice Date is a header-level attribute, and corrections are learned at the invoice header level, not the line level.

F. Invoice Number is a key header-level attribute, not a line-level attribute.

G. Invoice Requestor is a routing attribute, not a standard line-level attribute extracted from the invoice image for learning.

1. Oracle Fusion Cloud Financials

What's New in Oracle Cloud: Financials 21D

Document ID F45903-01

December 2021

"Payables" section

"Adaptive Learning for Invoice Lines" feature description. The document states

"When you make corrections to the following invoice line-level attributes in the interactive invoice viewer

the application learns from them to improve recognition for future invoices: Amount

Quantity

Unit Price

UOM

Description."